F Define and understand relevant
data. Know where the data come
from, how old the data are, and
what they measure.
Some publicly available data,
such as that from the Centers
for Medicare & Medicaid Services’ (CMS’) Five-Star Quality
Rating System, are used to
compare SNFs. However, CMS’
re-hospitalization metric is based
on hospital claims data, not SNF
data (see “A Level Playing Field”).
F Insist on data quality. Data quality is important in predictive data
Five-Star Quality Rating System data sources are often two
or more years old. Additionally,
because they are based on hospital claims data, they may not be
clinically accurate. Often, a hospital billing department adjusts
codes on a claim to ensure
reimbursement from a payer. As a
result, codes may not accurately
reflect the full complexity of the
patient’s clinical diagnosis.
F Focus on actionable areas.
Focus on areas that achieve
change within care facilities.
Emphasize the clinically specific
data related to re-hospitalization
rates when determining which data
to use in the predictive approach.
F Communicate findings using
transparent data sources. Share
data with each provider, whether
an ACO, HMO, or SNF.
Collaboration with various
care providers is necessary
to deliver quality health care
and implement integrated care.
Transparency of the data source
is instrumental when it comes to
While these are best practices,
some care providers and SNFs may
be reluctant to share facility-level
data, especially in today’s competitive marketplace. Some may
be reluctant to compare patient
outcome data and participate
in total transparency with other
organizations. Yet, when silos
break down between care settings
and a high degree of coordination
takes place, the result is improved
Culture of Due Diligence
The entire healthcare system suffers from long-term and post-acute
care challenges, including an aging
populace, a shortage of primary
care physicians, and difficulty
in knowing how to plan for the
system’s future. ACOs and HMOs
emphasize value-based reimbursement to assist in delivering an
efficient Medicare system. Big data
and data analytics are designed
to meet these challenges, but all
data are not uniform, nor are they
transparent or trustworthy.
assessment of each patient is
essential to person-centered care.
The MDS does this, but information
must be accurate to be effective.
In the best cases, the SNF’s MDS
coordinator and interdisciplinary
team collaborate on frequent
reviews of relevant, actionable
MDS assessment data.
SNFs may be technically compliant in their MDS regulatory
requirements, but to have a
true culture of due diligence, all
staff must insist on providing
high-quality data. This is accomplished by embedding a process of
self-auditing and submission compliance. An additional component
is communicating findings using
transparent data sources that caregivers and staff understand. In this
way, the process provides a true
picture of the SNF’s performance.
Quality, affordable health care is
a vast issue that requires everyone’s attention and participation
to create viable solutions. A level
playing field will only come from a
standard set of information generated from patient assessment and
care records. Using predictive data
analytics derived from standardized data is a part of the answer.
Transparent and free data-source
sharing provides a risk-adjusted
measurement of performance for
comparing SNFs and making care
coordination decisions beyond the
raw re-hospitalization rate and Five-Star Quality Rating System. In due
course, this will provide high-quality
health care, reduce costs, and minimize re-hospitalization rates.
Jennifer Gross, B.S.N., RN- BC,
RAC-CT, CPHIMS, is senior healthcare specialist for PointRight, Inc.,
which provides data analytic services
nationwide to skilled nursing facilities,
hospital systems, and post-acute care
clients. PointRight is based in Cambridge, MA.
1. Healthcare Information and Management Systems
Society. 2016. Care Coordination and Analytics:
Hospitals and LTPAC Data-Driven Partnerships.
HIMSS LTPAC Roundtable, December 14, 2016.
Accessed August 6, 2018 at himss.org/getinvolved/
2. Centers for Medicare & Medicaid Services 2017.
Identifiable Data Files. Accessed August 6, 2018 at
A predictive approach
requires more than just
collecting data. It provides
the structure and metrics
for tying together patients’
healthcare outcomes to
measure the performance
of SNFs. Policies can then
be enacted based on these
findings to drive appropriate